Pattern Recognition on NBA Halftimes

[TheNeuralNBANet]

I have a PhD in Artificial Intelligence. Trying my hand at sports handicapping. Been a casual bettor for years. in 2011 backtested a system based on last 1/3 of season and had computer instruct what were the most important variables (of 12+ available) for predicting 2H outcome. The result is a 58% strategy that picks alot of games. I will post them here to gain credibility/following for next year. Can't guarantee I will post them all but will be honest and open with picks and records for the days I post. As long time bettors know there is no such thing as a lock, only edges, small or large. A large edge would be 10-15% in my (mathematical) opinion.
Will add to this thread during season.
Dec record. 0-0
YTD record 0-0

 

[r.a.]

Welcome and appreciate your contribution ... bol.

 

[TheNeuralNBANet]

I bet this last year for myself and was 240-171. (I don't expect you to believe that with all the salesmen in this, but just as background)

 

[TheNeuralNBANet]

Ind/Port over 96

 

[Donkwin47]

Welcome to The RX and best of luck to you!

 

[candylicker]

Nice call

 

[scrubs]

I have a PhD in Artificial Intelligence. Trying my hand at sports handicapping. Been a casual bettor for years. in 2011 backtested a system based on last 1/3 of season and had computer instruct what were the most important variables (of 12+ available) for predicting 2H outcome. The result is a 58% strategy that picks alot of games. I will post them here to gain credibility/following for next year. Can't guarantee I will post them all but will be honest and open with picks and records for the days I post. As long time bettors know there is no such thing as a lock, only edges, small or large. A large edge would be 10-15% in my (mathematical) opinion.
Will add to this thread during season.
Dec record. 0-0
YTD record 0-0

does the model take into account injuries and changes to a team's normal lineup?

 

[TheNeuralNBANet]

does the model take into account injuries and changes to a team's normal lineup?

Not as such, but I'm working on that. I can explain why thats not such a huge deal but would rather keep it black box at this point (well, forever actually). Suffice it to say if e.g. LeBron James was suddenly hurt I wouldn't play Heat over or under for 2-3 games.

Dec 1-0
YTD 1-0

 

[deryckp]

thanks Neural. half times are hard to catch and wager on.

 

[Murphy8276]

thanks Neural. half times are hard to catch and wager on.

just catch the plays if he posts them here. we do not allow outside links anyway so please do not ask for them.


GL guys


-murph

 

[TheNeuralNBANet]

thanks Neural. half times are hard to catch and wager on.

Yes,
A) there is limited time window and
B) there are low maximum bets and
C) the line moves quickly and
D) a 1 pt move can make difference between playing or not (in long run 1 pt equals difference between -110 and -128) but
E) higher win rate can offset these handicaps, in my opinion. lotta randomness tho (as there is in all sports)

 

[RussellPeters]

Welcome and good luck!

Miami down by 11 at home to the Pistons as a 10 point chalk. Your system have any analysis on this 2nd half play?

 

[TheNeuralNBANet]

The under is 1/2 pt away from being a play. But no play (of late miami come storming back, score alot and go over)

 

[TheNeuralNBANet]

Tor/GS Under 102.5

 

[PhinFan89]

Nice

 

[TheNeuralNBANet]

Dec 3 1-0
Dec 2-0
Ytd 2-0 (usually more plays than this)

 

[BRICK1]

Nice start!

 

[TheNeuralNBANet]

Nice start!

2-0 is better than 0-2, but with any sporting event there is a great amount of randomness. I like to think of my algorithm as giving me the highest edge I could conceivably get (using maths), and maybe can improve it a little over time, but still that means I might win 60% and lose 40% if everything goes well. For anyone following this system or ANY system you really need 50+ plays minimum just to have an opinion. A 60% system can start easily start 4-10 and a crappy system can start 8-0. Most in sports (in my opinion) don't understand this - they fall prey to the recency effect - follow recent winners, stay away from recent losers etc. Best imo is to find something you believe in, use modified kelly to suit your risk profile and grind it out. But yes, I'll take 2-0.<><>

 

[Defying]

interesting. looking forward to seeing this in action!

 

[scrubs]

Welcome and good luck!

Miami down by 11 at home to the Pistons as a 10 point chalk. Your system have any analysis on this 2nd half play?

....please let's NOT turn this thread into the Bookie Report thread where you and everyone else starts asking the original poster for opinions on every game on the board.

he is posting what are plays, leave him alone to do just that and be thankful for it, win or lose..