Contest Idea

[j]

I have to start by saying I don´t want to grade this thing, but I had an idea I thought might be interesting to talk about even if it never gets passed the idea stage.

How would we run a contest on taking NFL leads?

What I´m thinking is that someone would post the opening numbers at say 11:00 PM EST Sunday, and then you´d have to have your ¨pick¨´ in by 700 AM the next morning. For every half point the line moved in your favor, you´d get a half point, and against would count against you.

The tech question, of course, is how to give the contestant more credit for picking the move from 2.5 to 3.5 than from 10.5 to 12.5, but I know there are folks here who could provide a multiplier chart that would make that problem a snap.

Since lead-taking is a subset of middling I thought it might be of interest. Of course, analysis and discussion of why a game might move (and then why it does or doesn´t move) would make the contest that much more valuable.

[uncle goopy]

Good idea. Don't think there'll be so many players that grading will be a problem, so I'll do it. Will also offer the canned ham I won in last year's UDog Classic as the prize. If j is willing to put up his ham as well (from the same contest except 2 years earlier), that would be quite a haul for the lucky winner.

Will bring the idea up again just before week 1 lines come out. Might be a good idea to move the time back until the early line differences between shops are hammered out -- opinions on that?

[ChuckyTheGoat]

I like it. But don't you guys think this is more applicable toward college FB gms? Could be wrong here, but I think coll lines move more than NFL lines. NFL opening #s by and large are pretty sharp and move less. I could be dead wrong about that.

[The Actuary]

It sounds interesting J, I found the few "set the spread" posts regarding the WNBA numbers before they came out to be extremely informative.

[j]

Chucky,

I know college lines are more volatile, so you´re definitely right a college division of the contest would make a lot of sense. My only reason for suggesting the pros is that I´m totally clueless about college, where as I´m just not as informed as I´d like to be about pros.

Also I know the count and others have access to some good conversion charts for the value of a point relative to a possible middle, and I thought that hammering out just how to grade the contest might be an interesting tech talk topic.

I guess too that middlers tend to look for juice deals to close there deals, and I think these are a little more findable in the NFL.´

[deepsouth]

I absolutely love this idea. I am not planning on wagering or really even paying attention to the 2002 football season, so I will likely get destroyed in this contest if it ever does come about ... but it's a great idea nevertheless.

I doubt we could get anyone to sponsor such a contest because a) there likely won't be very many participants and b) unlike other contests, it is highly likely that the winner of this would be a pretty sharp player.

I'm unsure of how to work the logistics of this, what with the differences in lines between books and so on ... but put me down as one vote strongly in favor of a lead-taking contest.

[GottaWinToday]

Sounds like fun.........I'm in as a participant. Those OLD Hams, wow, what a prize.

GottaWinToday

[j]

Well cool...it looks like there´s small but very definite interest in this.

Anybody have any thoughts on grading it? That is, how do we give the guy who sees 2.5 going to 3 more credit than someone who sees 10.5 go to 11.5? Also, I guess this has to be a contest where you can actually lose points.

I suggest the name the ¨Junior44 Pick the Move¨ contest in honor of one of the great contest moderators of all time who has apparently left the posting planet.

Sorry I can´t put up my ham. I was going bad last year and I had to eat it.

[count zero]

Actually, I was thinking we should have a Junior44 Memorial Underdog Classic here at MW. Unless they're carrying on over at UC under a different guy.

Anyway, on the line move contest -- we need to start a line move thread, and then the params for the contest should emerge from what that turns up. Might be a good idea to put a link to the line moves thread in the Mess Hall -- a lot of the experienced guys there who don't come here might be able to add something.
Or maybe we should hash out a possible contest scenario here first, then run it by those guys for adjustment purposes.

Possible problems:

1) What to use for opening line. You can see the official openers at a variety of places, but they tend to get hammered out against each other almost immediately. There's no skill in seeing which way those moves will go. Once a consensus line is established, then there's a slowish adjustment up through the weekend, then some games move a lot as the kickoff nears (guru releases?). I think the move between Monday and Friday is the one you could make pay off if you had some ability to predict it. Don't really have any reasons for that, and am open to other opinions. But just for the sake of argument, if we decided to play on that basis, what would we use for the Monday and Friday lines? The line at Oly at midnight EST, something like that?

2) Valuing the moves. As j said, a half from 2.5 to 3 is a much better prediction than a half from 5 to 5.5. I think we have the talent in here to put a number on the value of each half point from 0 to, say, 14. Whether anyone will volunteer, or even take a stab, is another question. But if such a valuation could be established, then scoring would be easy.

[Machiavelli]

I think this a great idea for a contest. count zero has identified the two main problems and I will try to provide solutions.

(1) Which to use as an opener? I would think it would be important to have some type of median or consensus line. We know that Olympic and CRIS will be available Sunday night, I think we just need a third out and then we could take the median line from those three.

(2) Grading. I think it is pretty simple once we can get a number value chart. If the value of 5 is worth 1.6% and 6 is worth 2.4%, then a move from 4.5 to 6 would be worth 1.6 + 2.4/2 = 2.8 "pts" in the contest. Now, getting someone to share his chart might be an issue, so here is my suggestion: round all of the percentages to the nearest even whole number. I don't think the coarseness of the chart will hurt the contest at all. I think we could put it to a popular vote if people think that 10 should be worth 6% instead of 4% or whatever. This would simplify the scoring, too, as it would get rid of the decimals. Here would be my chart off the top of my head, but anyone else's estimate would be as good or better than mine.

1 = 2%
2= 2%
3 = 10%
4 = 4%
5 = 2%
6 = 2%
7 = 6%
8 = 2%
9 = 2%
10 = 4%
11 = 2%
12 = 2%
13 = 2%
14 = 4%
15 = 2%
16 = 2%
17 = 4%

[count zero]

Great, we're in business. I haven't updated my own chart for a few years, but yours certainly looks familiar. It'll obviously do the trick. I'll update my numbers and IM them to you, thanks for posting yours. I do still see a problem with establishing an opening line, though. Using a median of three early hangers is fine, but there's still some ambiguity. For example: Let's say the Giants are playing the Jets and books 1 and 2 both put up NYJ -4. A half-hour later, book 3 comes in at Jets -4.5, but in the meantime book 1 has moved to Jets -5 and book 2 has moved to Jets -4.5. There are a couple of issues to worry about here. Let's say we solve the obvious one by strictly following the rule Mach mentioned: use the median of the 3 opening lines. Book 1 = 4, book 2 = 4, and book 3 = 4.5, so the median would be 4. But now there's no mystery which way the move is going to go, because it's already under way. Also, we need to address the problem of many opening lines moving in response to arbitrage. Obviously, if there's a disparity between several sets of lines, the line at any given book is likely to move toward the line at books with a different number. It doesn't seem to me that this is what we're looking to reward in the contest.

Any thoughts?

[Machiavelli]

Good points. Stuff like that is what i was trying to avoid with the median, but it is impossible to freeze the time axis so that everyone can have a chance to enter their picks.

There does appear to be window where Olympic and CRIS are not moving their lines, but it is small. From 11 pm EST Sunday night to 10 am EST Monday morning, Olympic either didn't move their college lines or didn't report the moves to Don Best. CRIS had almost no moves in this timespan, but they did in fact move Wake Forest from 2.5 to 3 at 2 AM and UNC from 9.5 to 9 at 8:30 AM. I would imagine the same "cease fire" time range would apply for pros (?)

BTW, this is just a very coarse, off the top of my head estimate for the number values.

[The Actuary]

Post a line, give people a fixed amount of time to reply, higher or lower Bob. Winner is the most correct at the end of the year, I would suspect the magnitude and the location along the point spectrum of the move would wash out in the end anyway.

[count zero]

The more I think about it, the more I see all the problems stemming from nailing down a fixed opening line for us to shoot at. TA's idea, for example, would mean we'd all have to shape up at a certain time -- a time when I'm just about ready to pass out, after all the stress of watching the games, doing many celebratory bong hits, downing many beers to kill the pain, enduring the hectic business of updating my lines for the next week, etc. Is it central to the concept that we catch the early hammering-out phase? I guess that's where the sharps do their thing, so probably it is. Just seems it would be a lot simpler if we started with a more settled consensus number on Monday.

[The Actuary]

Are you Hunter Thompson or just his attorney?

[alysheba88]

Is the idea to pick lines or games?

The academic world has run amok

[cash]

aly:

It's a contest on anticipating line movement
A very good idea..give the edge to the
scalpers who are good at taking leads

[alysheba88]

Thats what I though. Ill stay out of that one.

Its like having a contest on the final odds for a horse race (and not the race itself). Thats the way I am reading it anyway.

[j]

Thinking tactically about this contest it already looks to me like the winner will be somebody who hits a few moves at the number of numbers, although, one of the good things about having a small number of participants is that it will give somebody consistently catches slight adjustments a decent chance.