Hooligans Sportsbook

Trouble - need help

  • Start date •
  • Replies
    115 Replies •
  • Views 7,905 Views
Follow
MrX
If you are able to hit > 60% of 100 games a season at -110 going forward, the only thing separating you from a pro is bet-sizing and the decision to widen your selection criteria. Count me among the very skeptical that your posted win% is anywhere near sustainable, though.

If you ever want, though, and if you'd be willing to divulge your methods in private, I can back-test your system.
Click to expand...


Mr.X, every play I have done the past two seasons are documented at Cappersmall and SBR. Pretty much all my threads are titled something like Wal's NCAA or College Football Plays. There may be a few variations as I was trying to find an identifier but something like that for titles.

The only secret I keep for myself is my personal factor/variable. Everything else I have devulged many times. I start with Sargin's as a base add my factor/variable (which is a sliding scale based on specific situations, teams and conditions), then the games that meet a pre-determined criteria are further evaluated through regular number and stat research as well as injury reports and such.

I post my adjusted lines each week. I didn't in the begining but started last season because people wanted to see them and many used them for themselves. All plays are posted as well. So anyone can fact check my results and varify anything they want.

If I took the time to learn how to import data through Excel things would be much smoother and I probably could better select my final options but I just haven't taken the time to figure out how yet.

Next week will be the first week for this season that I post "Program" plays and I'm confident that through the end of the regular season it will again hit over 60%. I've tried it in Bowl Season and other sports but just haven't had any success with it. The problem for other sports is in me failing to do enough side research into games. The problem with bowl games is there's just noway to factor motivation except for maybe the bigger games.
 
You reacted kind of like you thought MrX was doubting your claims about recent season results Wally. I don't think that's the case. I think he was offering to back check your system/model by plugging it in against previous years' stats/results to see how it would have made out longer term.

Perhaps that is not possible considering your personal variable/factor.
 
Muddy, I wasn't really sure if there was doubt but understood why there would be, that was the only reason I refered to where and how it could be checked.
Mr.X, if you thought I was offended or defensive, I wasn't.

As for backchecking the overall system, I don't know how that would be possible since I have crashed this computer since football season and I didn't backup the spreadsheets. Fortunately the way I determine my lines nothing really was lost except past data.

This year I have a backup harddrive and will begin compiling.

Mr.X or anyone for that matter that wants to talk shop but feel it best to do it privately you can @live.com me or FaceBook.
 
You only need an hypothesis and/or a set of criteria to backtest an approach Wally. It sounds like you're saying you don't have a defined set of criteria. If you do it's only a matter of feeding it all to MrX's robot so it can give a thumbs up or down to your special recipe. You don't need actual data.
 
It's not a trust thing Matty. Really it has more to do with the randomness of the variables applied. If I miss something then it is skewed but if I missed it good or bad is the part that can't be answered or even evaluated.
 
I had plumbing issues to deal with for a while there. I wasn't intentionally avoiding this thread.

I have the data and historical lines, which is why it's generally pretty easy to back-test someone's system. Although, I forgot that you start with Sagarin ratings. I think weekly Sagarin ratings are out there somewhere, but I don't have them.

So anyway, no, I wasn't really addressing whether your reported record was accurate, but since you bring it up, I'll give you my thoughts. There are a lot of issues with only forward-testing a system/model/whatever. Even if we completely eliminate any intentional fudging of numbers, human nature almost inevitably leads to some misleading reporting. One thing that immediately stands out to me is you stating that your model doesn't work during bowl season. I get the impression, then, that results from bowl season aren't included in your record. What if bowl season had hit at 80%? Is it possible that it would then be part of your system and would be included in your record?

I'm curious, too, about the first couple weeks of the season. Did you, right from the start, assume that your system would not work for the first two weeks, or did you have bad results and then come to that conclusion?
 
Good questions Mr.X and I do solove to type.

The season record is just for the regular season. I list a seperate record for Bowl games. Now in the mix are a few added plays such as late reverse line movement and 2nd half plays. I made a conscious effort last season to not include these types of situations but not from the start so yes the results could be influenced but the influence would only improve the win % as I make bad judgments interpreting RLM sometimes and less than 50/50 on half time plays.

As for starting from the begining of the season it's kind of muddy there. The previous 2 seasons I started from the begining but last season as I mentioned I had a little trouble because I tweaked it some and that didn't work so well. Now admiting the dumbass part of this. I realy never realized until I was looking at the Sargin's data that it was previous year numbers to begin with. OBVIOUS as that would be to everyone looking I simply wasn't even thinking about it. I did realize it this season and also understaood that not every team was playing the opening week. That is why I have stated that the program plays will begin Week #3. Teams will have played a game and updated ranking will have been adjusted. I also realized this is why Weeks #5-#10 are much higher win rates. Things at that point have settled out and the Sargin's is better qualified to estimate the power ranking of the teams.

For years I toyed around with compiling and computing my own ratings system, I would hit a good couple of weeks and then it would all go wrong. I adjusted and tweaked but nothing really seemed to work. Sargin's changed all that. Whatever parameters go into creating those ratings just works. By using Sargin's as a base and expanding on it just works for me.

Now I admit the biggest flaw or maybe the best attribute is my variable. It's flawed because it's random but it works. It's an attribute because just using the Sargin's really isn't much better than 50%, sure it fluctuates but it'll be about 50% on the average.

This year I am adding more dicipline. I'm not doing add-on games, I'm keeping my cards relatively small (not that I have ever had more than 11 games in any 1 week) and I'm sticking to my philosophy of treating all games equally as far as wagers go. I always say this but then I usually get all random with the bets.
 
I'm not totally clear on how all of that affects your record-keeping, but the bottom line is if you're playing games (bowl games, 1st two weeks, added plays, etc.) that could get added to your "official record" if they go well, but end up being dropped if they're unsuccessful, your numbers are not valid.

Even if you've steered clear of all of these things, you should still expect your win% to go down. There's a lot of luck in samples of this size. You mentioned that you've tried this approach on other sports without success. You came to the conclusion that you just don't analyze those sports as well, but there's actually a really good chance that if you did it all over again, you'd consider NBA to be your best sport.

I'm not saying any of this to bust on you or to wish bad luck on you, just letting you know what it looks like from an outside perspective.

Is there any way you can explain the "random" nature of your "personal" variable? I'm at a complete loss as to what that means.
 
mexican-travel-advisory.jpg


krayz-300.jpg


custom_1238094285892_Photo_114.jpg



Random enough for ya? PAY ME!
 
Ok, explination of randomness.

You know how it is said that a team has a 3 point edge for homefield advantage, well that may be true for some teams but not every team. I have my own variables for homefields, level of intensity for certain situations so forth and so on. I call it random because it's not the same for every team and not the same for every situation. To add to the randomness it could change for the same team, on the same field but playing under different circumstances.

So as I've said this variable is flawed due to it's randomness but at the same time it has been they key to making what I have done work.
 
You must log in or register to reply here.