[Dark Horse]

Looking at BTP results. Only considering records of 12 weeks or more, at least 48 picks.

Last season 129 contestants played 12 weeks or more, and included at least 48 picks:
11 hit 60% or better.
21 hit 58% or better.
28 hit 57% or better.
38 hit 55% or better.

This season, with 278 players (!) playing 12 weeks or more, only:
12 hit 60% or better
22 or 23 hit 58% or better (1 game left)
35 hit 57% or better (1 game left)


Consistency. From last year's group how did the players who picked over 55% last season do this season? First percentage is last season, second percentage is this year. If you haven't seen this type of difference before, it's pretty shocking... (if there's only one number, the person either didn't play this year, didn't enter 48 picks, or was missed by yours truly).

 jake38 67.3%
gerry 64.8% - 42.9%
aakj 64.3% - 55.0%
rta21 64.3% -37.9%
Sinister Cat 63.3% -47.1%
SexyMit 63.6% -39.7%
Boner_ 63.0%
losturmarbles 62.7% - 45.0%
Iceman 62.0% -57.6%
sharktank1 60.6%
Mudcat 60.0% - 50.0%
SilverFox8 59.3%
vbomber 59.3%
ertl09 59.3% -55.3%
Doc JS 59.2% -44.1%
inspades 58.8%
Hoja Verdes 58.7%
Ironhorses74 58.3%
pats3peat 58.2% - 50.0%
capitalist pig 58.2%
Dark Horse 58.2% - 52.6%
Willie Bee 57.7% - 41.2%
alta 57.7% - 58.6%
mr x 57.4%
Doug 57.1% -52.5%
Poker_Beast 57.1% -40.7%
rjt721 57.1% -45.5%
blackbart 57.1% - 42.9%
madworld 56.6% -50%
McBa1n 56.6% - 41.7%
Order66 56.4% - 31.5%
linebacker 56.4% -49.1%
hawk 5 56.4% - 47.5%
hhsilver 55.8% - 51.9%
ldrapeau 55.8%
jon13009 55.6% -52.5%
clarkd32 55.4% -50.8%
iwinyourmoney 55.4% -59.3%

This is a two-season comparison, and isn't meant to offer a complete picture. If I do this again next year, I will include those who picked 55% or higher this season. On this list only 9 players averaged 55% or higher over two seasons. For the 27 players with two qualifying records, the average drop in percentage from the previous season was a shocking 11%. The average drop for players hitting 60% or higher in the previous season was 16%! Players that hit from 55 to 59.9% dropped by an average of 9%. Books love these type of fluctuations. Kelly bettors not so much.

[RickySteve]

Are you saying I should stop paying gerry $5995 a year for his plays?

[Dark Horse]

Too late now, but fading those plays was well worth the money. The average for the players that hit 60% or more in 2008 was only 47%. That's for a sample size of roughly 400 plays in each season. Take out sharps like aakj and the number drops even more dramatically. Pretty amazing.

It also suggests that books should encourage winning players to bet more rather than limit them; and never limit until they have a few years on record, instead of a few months.

[durito]

Fading a 50% player when they are cold isn't any different than following them when they are hot. It's a negative proposition.

Any intelligent book is limiting based on betting patterns, not $ won anyway.

[Iceman]

Best NFL contest going, IMO.

Thanks SBR!!!

[dwaechte]

Too late now, but fading those plays was well worth the money. The average for the players that hit 60% or more in 2008 was only 47%. That's for a sample size of roughly 400 plays in each season. Take out sharps like aakj and the number drops even more dramatically. Pretty amazing.

It also suggests that books should encourage winning players to bet more rather than limit them; and never limit until they have a few years on record, instead of a few months.

I hope you're kidding.

[dwaechte]

How did you find last seasons standings anyways? I'm interested in seeing the full standings for 2007 and 2008 if anyone could find them. I've been searching around and it seems as if they've disappeared.

[Dark Horse]

How did you find last seasons standings anyways? I'm interested in seeing the full standings for 2007 and 2008 if anyone could find them. I've been searching around and it seems as if they've disappeared.

I can't find the link to 2008 right now. Will post it if I find it. I have the standings, but wouldn't look too good to copy and paste. This is you in 2008:
 dwaechte 10 $271.28 9 12 25 21 2

[Dark Horse]

I hope you're kidding.

Comparing consistency over several seasons gives a better understanding of how much is skill and how much is luck. I would not fade a skilled player, but if a lucky player is just off a major upswing I might consider it. I wouldn't do it, because I need more information for a bet, but that's personal preference. If the next season shows similar results I would be willing to bet that the majority of lucky players (identified by huge swings) who hit over 60% one season will not hit over 50% in the next season. NFL only, because of the limited number of games.

Most people probably see upswings and downswings as isolated from each other, but I see them as cyclical (part of the same cycle). To me it's the same difference as between a linear view of time and a cyclical view. I understand that this might by unacceptably vague to math minds, but they forget that their entire 'universe' is comprised of laws and formulas that others discovered for them. Before these were discovered they too would have been vaguely defined (but would have worked just as well in Nature). So math minds that laugh at ideas beyond their own world forget the basic principle that defined that world. That of investigating the unknown, that of looking at the same things in new and unexpected ways.

Those who embrace the unknown typically are inspired by their own lack of knowing, or ignorance, as a driving force. Those who shun the unknown often replace their own ignorance with arrogance, a false sense of superiority. It is very difficult for these two groups to get anything across to the other.

[aggieshawn]

nice info

[SexyMit]

Very good info. The line movements that I followed last yr, went completely the opposite this year. But has worked for quite a few years until this year. The Public won alot more this year then last year also, books took a beating a few weeks the 1st couple of weeks. Oh well Football didn't fare so well but the Nba is treating me right!

[dwaechte]

Very good info. The line movements that I followed last yr, went completely the opposite this year. But has worked for quite a few years until this year. The Public won alot more this year then last year also, books took a beating a few weeks the 1st couple of weeks. Oh well Football didn't fare so well but the Nba is treating me right!

Go tout your mom.

[reno cool]

Haven't you considered variance or regression? I mean I would think it's pretty fair to assume the high hitters from any year simply got lucky. Plus consider the notion that some would quit if they got off to a poor start and not meet your minimum requirements. The pricks as a whole hit about 48.5% this season, (plus I believe the avg player hits less than 50%) . Combine that with a relatively small sample size and possibly multiple bets on the same side I don't think there's anything surprising in the results.

[Dark Horse]

Haven't you considered variance or regression? I mean I would think it's pretty fair to assume the high hitters from any year simply got lucky.

Exactly like blindly picking balls from the proverbial bag with 450 white balls and 550 black balls. If you're on pace for picking 650 black balls early on things will even out later on. One season is clearly not enough to complete that process. So one question is: what is a reliable sample size - how many separate seasons - for the NFL (and NCAAF)? And how fast do things regress? I was certainly surprised by the apparent speed of things evening out, as in how often a sharp upswing from the previous season was followed by a sharp downswing the next season. Maybe that's coincidental. Next year's numbers will offer more depth.

I'm mostly interested in the BTP breakdown to study the longer term differences between skill and luck. When you only look at your own percentages it's easy to miss a bigger picture. With this type of info you can place your own numbers in a larger context.

[SBR_John]

iwinyourmoney 55.4% -59.3%
Iceman 62.0% -57.6%

....and of course
Dark Horse 58.2% - 52.6%

Great job by Iceman! Thanks Dark Horse, excellent info.

[seaborneq]

I am glad you didn't use my percentage. I won good money last year. Only 1 winning week this year. I would ruin the curve.

[btraband]

good research, I always believe in hot streaks and cold streaks but still average around 50%.

[Boner_18]

Wow good info. I went less than 50% this year i believe. Did much less capping and less weeks b/c of the number of people sharing prizes and the book requirements.

[Mudcat]

Mudcat 60.0% - 50.0%

Not to get too deep into this but I play different when I am in contention for the big money than not. I'm probably not the only one.

In the years when I was hanging around the top 10 in BTP, I take it seriously and enter my true best picks. When I got off to such a bad start this year - my Best Bets did badly and had me in negative units - a lot of times I would just pick a TV game or a big favorite like New Orleans that I wouldn't actually bet money on, just because they are fun to watch and it would give me another game to follow.



If I knew a list like this would be coming, my vanity might have kicked in and I might have played it more straight up. Well probably not. But my actual figures have been quite consistent the last two years.

[blackbart]

you would not expect the results to have a normal distribution with such a small sample size, simply due to variance. if you analyze the sample as a whole it will be normal i am guessing

[Dark Horse]

Same changes in motivation can easily apply during regular (non-contest) season, and still don't explain shifts from 60 to 40's. Less interest only increases the monkey-flipping-coin element, said to even out at 50/50. I wasn't planning to make this list in this form; and I might have tried harder as well. The past season seemed like slippery ice to me, compared to seasons where one has 'traction' throughout. In any case, I know that pros can easily make money during a 50% season, so didn't think it a big deal, **. To me it's much more interesting to see how much people can control their bottomline, and not fall lower, during a season without breaks, than how high they can fly during a season with tons of luck. And **, you're also one of a few hitting at least 55% over past two seasons.

The switches in this list came as a surprise to me. Quite interesting that more than a few here think they mean nothing. If it was one or two switches from 60's to 40's I might agree, but in this case they were the majority.

[SBR_John]

Its too small of a sample size. I suggested going 5 picks over 20 weeks last year. I still think we need at least 100 picks. Or maybe even make BTP a yearly thing with 200 picks/40 weeks and double the prizes/cash across the board??

[Dark Horse]

The sample size will increase next year. It's not that small. You're talking at least 96 picks for each individual, so multiply (roughly) by number of people involved for the group sample size.

[MilfDriller]

Order66 56.4% - 31.5%
linebacker 56.4% -49.1%
hawk 5 56.4% - 47.5%
hhsilver 55.8% - 51.9%
ldrapeau 55.8%
jon13009 55.6% -52.5%
clarkd32 55.4% -50.8%
iwinyourmoney 55.4% -59.3%
durito 49.2% - 47.5%



there you go. you forgot one.

[Dark Horse]

Please don't take this stuff personal. Like losthismarbles. It's a group sample intended to look at general trends. Not an individual sample to look at individual trends.

If there is no general trend, fine. But I'm not willing to assume that after seeing these results. I now want to know if there's more to it. That's all.

[KKoz9]

Did anyone ever find the links for 2007 and 2008?

[marcoforte]

Your data proves why kelly betting is difficult because you need to have some idea of what the future win rate will be based on a prior win rate. Very few follow through in multiple years.

[Dark Horse]

Its too small of a sample size. I suggested going 5 picks over 20 weeks last year. I still think we need at least 100 picks. Or maybe even make BTP a yearly thing with 200 picks/40 weeks and double the prizes/cash across the board??

Brass doesn't listen to you anymore?

Would be very cool to have some form of BTP year round, especially if others can win the role of Prick outside of football season.


FWIW, in my experience 7 NFL picks per week is best. With 4 picks the short term fluctuations are sharper, and with 10 picks there aren't enough strong plays.

[Chi_archie]

great thread

[jon13009]

A winning % over 52% is a tough thing to produce in any year.
Over a 3 year period it would be even more difficult to produce consistent results.

Interesting. Thanks.

I have an image of BTP standings from 2007-8 (my first and personal best year at BTP) That year was golden because SBR actually mailed you a valid check for that weeks winnings without the hassle of a 4x rollover.

The only thing consistent about the BTP contest is that year after year the requirements of this contest get more complicated each year. This year SBR screwed a lot of people (including myself) by requiring you to sign up for 16 book accounts and if the account numbers you gave SBR got screwed up you got screwed.

Perhaps SBR could release a spreadsheet of past winners and their percentages?

[Chi_archie]

very interesting