NHL Picks, Sat 1/25. This season: + $13 541 (250-203-45)

[Mathemagician]

Today’s investments:

Calgary ML +130 / estimate +108 / wager $300
Carolina ML +115 / estimate -112 / wager $300
Columbus ML -105 / estimate -120 / wager $300
Ottawa ML -270 / estimate -385 / wager $300
San Jose -0.5 +115 / estimate +102 / wager $300
Montreal -0.5 +120 / estimate +103 / wager $300

Nsh - TB Under 5 +115 / estimate -106 / wager $100
Bos - Phi Under 5 +117 / estimate -112 / wager $100

Click here for Mathemagician’s detailed track record and more information for NHL 2002-2003

-----------------------
Mathemagician’s analysis is based on many years experience of odds making in NHL and other sports. Methods used in odds making process include careful pre-season analysis, continuous measuring of each teams abilities using power ratings and considering the daily news in each individual game.

The pre-season analysis includes mathematical and statistical analysis and determination of the changes caused by the transactions. The power rating database is always updated after each game and every published estimate is adjusted to consider injuries, fatigue caused by the schedule, different motivations and many other factors.

The pick is profitable in the long run only if the current price is higher than the estimate. Otherwise the bet is not worth placing.

 

[ScoobySnacks+]

Math,

First off congrats on the nice night!

Just a question though, awhile back I asked about your ROI and it being at app. 115%

Now I was just figuring mine out (hopefully the right way) and for the sake of this explanation I only used even $$ for wagers placed (I lay some wood once in awhile, but never too extreme)

Now my contest plays are at 354 for the year, saying a unit was 100$ that would be $35,400 wagered (at even or +$). Return on those wagers would be $5765.

5765 (return) divided by the 35,400 (investment) would equal around 16% ROI, is that not the proper way to figure it out??

I might be figuring it out wrong, that is why I was amazed at your over 110% ROI.

Thanks

& GL

 

[ScoobySnacks+]

OK I got this from your site...

"The ROI-column is Return on Investment using the flat stakes from American format picks. The ROI is simply all winnings divided by the sum of all stakes. And if the ROI is over 100% then the picks are making profit."

Not wanting to sound like an A-hole for asking this, but if you lay 300 (flat) on -300 you would win 100 back, a few of those would drop down your ROI a few % points even if you nailed everyone of them.

May be because its pretty late and I should be in bed fondling my wife, but 115% return now seems actually out of reach, unless your equation adds 100% to your total somehow (like you state about making a profit if its over), and your ROI is app. 15%??

Again, I'm just curious and don't want to twist your Joe Boxers

Time for some quality time with the wife

GL today!!

 

[Mathemagician]

The ROI in my records is calculated using the gross winnings, not the net wins.

So, if you lay 300 on -300, the win amount would 400 and ROI 133%.

If you have made $5765 profit after wagering $35,400 your gross winnings have been $41,165 and so the ROI is 116%.

 

[ScoobySnacks+]

Thanks Math! Hope you nail them all tonight!!

GL

 

[alwayswins]

I had followed your picks for awhile, but I was disappointed in the results so I stopped watching. I realize that short term results from too small a sample cannot be used to estimate the population distribution from which that sample was taken. However, because I have limited time, I have to cut time from activites that seem to be going towards dead ends. Your record since you began here is 41-41-10 (250-203-45 - 209-162-35). I don't say this disparagingly, because, as I stated above, the sample I saw was too small to judge someone's skills, I am writing to ask if there is a way to determine whether one of your picks is superior to another. On a day when I see 4 or more picks, I can understand that they all may be high probability choices. However, when I see many days with 4 or more picks, I can not accept that they are all high probability picks. Some must be low. If it can be determined, I am interested in the ones that have the greatest chance of winning, assuming that they can be identified.

 

[Mathemagician]

Our estimates indicate the probabilities to win.

For example, if our estimate is +108 it means that we think that the
probability to win with that pick is about 48%. (If the estimate is +100 or
greater you can convert it into a probability like this: 100/(108+100) =
0.48 = 48%)

And if our estimate is -385 we think the probability to win is about 79%.
(For estimates of -100 or less the convertion into a probability goes like
this: 1/(100/385+1) = 0.79 = 79%)


But I want to emphasize that the probability is NOT the criterion that makes
some picks better than others.

If you want to make profit you should look for value and not just the pick
with highest probability. Value bets can be identified by comparing the
price to the probability.

Let's use two of our picks as an example:

Carolina ML +115 / estimate -112 / wager $300
Columbus ML -105 / estimate -120 / wager $300

Here we estimated that Columbus is a little more probable winner than
Carolina is. But when you look at the prices you see that Carolina is paying
lot more and therefore it's a better pick than Columbus is.


Our record at the time of the post was indeed 41-41-10 but it's not the best
way to measure our success since it doesn't indicate how much we have made
profit. At that time period we had wagered $20 400 and our gross wins were
$20 953. The net profit was +553 USD. This is because we bet more on the
high priced underdogs than the low priced favourites and therefore we don't
need to pick that many winners in order to make profit. After Saturday's
picks the record is now 46-43-11 with +$1201 profit. But as you stated
yourself this is not enough bets to make serious conclusions. The full
season record gives much more reliable indication of the profitableness of
anyones picks.