NHL Picks for Monday. This season: + $12 897 (219-175-37)

[Mathemagician]

Edmonton ML +125, stake $300
Ottawa ML -154, stake $300

Min - LA Over 5 +130, stake $100
NYR - Ott Over 5.5 +115, stake $100


Click here for Mathemagician's track record for NHL

 

ott -.5 -10 is better than ott pk -54.
so why bet the pk price?

bb

 

[Mathemagician]

When using flat stakes, OTT -½ -110 is indeed a slightly better price in terms of expected profit.

But with variable size stakes (the Kelly Criterion) the difference in losing probability (almost 10%-units in this case) would make the moneyline a better pick.

For anyone who uses proportional stakes from the bankroll (instead of flat stakes), Kelly Criterion is a better criteria for choosing between moneyline and puckline prices.

My own (probability-) estimates for Ottawa are -302 (1.33 in decimal pricing) for moneyline and -196 (1.51 in decimals) for -½ handicap. And now that the bookmaker odds are 1.65 for ml and 1.91 for -½ in decimal format, the corresponding Kelly-stakes would be:

ML: (1.65/1.33-1)/(1.65-1) = 0.37
-½: (1.91/1.51-1)/(1.91-1) = 0.29

And from this we can see that the optimal moneyline stake is much higher, indicating it's the more profitable one in the long run.

Even if you don't bet according to the Kelly Criterion (and I don't recommend anyone to do that), this does prove which one is theoretically more profitable bet.

And since I don't know how much other people wager according to my picks, I can offer safer picks this way. It's up to the people to decide how how much they want to bet. I just post the best possible picks I know.

 

...just TELL people that ott -.5-10 is indeed better, then give them the option. for instance, i like minnesota today, but under no circumstances would i advise someone to lay -.5-35, so instead i send out min -1.0 even and then THAT gives them the option(s) to look for -1.0, -1.5+$, or props. in other words, why would you not post BOTH options? instead, by doing it the way you are, no one who bets seriously will EVER take you seriously. when you post big favorites (part. like the -300) not only will i be laughing, but so will everyone who bets.

bb

 

I like your picks. Any chance of you sharing your formula for capping hockey?

 

[Mathemagician]

Unfortunately I can't reveal my handicapping methods in detail.

But here's a question (and answer) from our website, which explains the general outlines on how our analysis is made:


Q: Are your picks computer generated or 'hand-made' by human experts?

A: Our picks are possible only because of the 'co-operation' between computers and human experts. If one of those two were missing, there would be no SportsDiscover.com picks.

Our analysis and research team consisting of top level mathematicians, computer experts, professional athletes and professional handicappers have created the state-of-the-art computer software for making sports analysis. All important aspects of the sports games are mathematically modelled in the form comprehensible to a computer. Simulated Artificial Neural Networks are even used to find the possible factors human experts aren't aware of. All mathematical models, methods and factors are tested and optimized using historical data from tens of thousands of games. Automatic web-spiders are used to collect all the new scores and lineup-information and input them into the system after each game.

Computers make such tasks reality which would be totally impossible for any human being to handle. No person could go through 100 000 games over and over again to find new angles and factors for use in sports handicapping. But with a computing power capable of doing millions of floating-point operations per second it's just a matter of minutes. The quality of computer analysis is also very solid. It never gets tired or try to do things in a hurry with the expence of quality. And the achieved quality level can never get worse. It can only get better when more data is gathered for the optimizations and also when the human experts are doing more research and testing.

Human touch is in a big role in a way that most of the mathematical models are developed by people with deep understanding of the sport. After that the computer takes over the routine work while human experts stay in the supervising role. The computer then outputs the analysis including all numerical data and the short standardized text-based explanations used in the website. This process makes it possible to do analysis by the human way of thinking boosted by computer's unbelievable computing power. All daily information on the web site are also automatically updated by the computer, minimizing the possibility of any human errors.

 

[zoom]

Wow, you must have pulled out the golden shovel to come up with that crap.

 

[Juggler]

MY Computer likes SAN JOSE. One of us has to reprogram. /infopop/emoticons/icon_smile.gif

 

[tuck321]

Bobby always bothering you is a fairly classic case of niche ecology so I would take it as a compliment. On todays games I am betting Rangers under. In the non contest real world I bet under minus 15 cents. My question is why you are using the silly lines for your picks instead of the easily available ones at Canbet and Pinn?

 

well if i wasnt confused before i sure am now. niche ecology?

bb

 

[Mathemagician]

For my picks I use lines that were very much real at the time I made my picks. The bookmakers I use are mainly the ones you mentioned: Canbet,
Pinnacle, and also Olympic. However, it is a fact that the smart money moves the lines so much, that at later times you don't get prices that good.

My estimates are available to everyone interested at my web site approximately 24h before the games start. Using those estimates you are able to make your own picks and get the best of the early lines.

 

i confess that my eyebrow was raised when you mentioned you got feedback from former professioanl athletes. since its common knowledge that players/coaches are worst handicappers around, i was curious exactly what they had to offer?

bb

 

[Mathemagician]

It's very true that they usually aren't very good handicappers (it's
possible though).
But what they have to offer is deeper understanding and different angles
about sports. Watching games and talking about sports generally with
professional athletes gives many ideas for research that non-professionals
rarely even think about.
So, it's definitely not the picks they might give

but the ideas to make
more research.

 

i use tons of numbers/stats/trends etc (often end up 'over-handicapping') here so anyone having a program which can do what i do by hand intrigues me. however i am naturally suspicious of anyone who maintains that laying prices is a good thing, software or no software. there is so much more to handicapping than a printout: reading between lines, watching (and know what one is seeing) games, evaluating injuries and their worth to number, having a general feel for the game. i just dont see how that can be programmed into a computer. and a computer certainly cant predict which way a line will move, which is def. a part of betting. does the computer tell you when its time to 'get off' a game? or buy some back? the whole concept seems limited to me.

bb

 

[Mathemagician]

Actually, predicting the direction which way the lines move during the day
is one of the easiest tasks. Compared to my own lines, the opening lines
from the bookmakers are (almost) always further off than their closing
lines. So, as a general rule the sooner you make your bets, the better price
you get.

You are right that there are things that our computers are incapable of
doing, like watching the games etc. But after processing all available
numerical data, the value of "personal feeling" is not that big. The Vegas'
NHL lines are suprisingly easy to beat as long as you do your math
(statistical work) well. (Notice! This doesn't mean that we would pick 70%
winners or anything like that, but it means that we have been able to make
profit in the long run after betting thousands and thousands of bets.)

The bookmakers try to post their lines according to the real probabilities
of the game (as well as to balance the betting action). And with a house
edge less than 5% it doesn't require big differences in opinions to find
value from one side or another. There is absolutely no reason why you could
never find value from laying prices. Bookmakers simply can't post lines that
would always be too obviusly biased for the favourite.

Still, majority of our picks are for the underdogs. That's where the value
often is, though not always.