Flat betting in Baseball....

Lawstyle, if my numbers are better because I bet slightly more dogs than favs, so be it, but I don't think this is the case. Do you have a formula I could use to find the average moneyline of games that I play? If so, I can tell you the answer immediately.

However, doubling units used does not change anything. It's like multiplying both sides of an equation by 1: it doesn't change anything. Just to double check that I am not losing my mind, I doubled all the units risked for both methods and I am still netting more units using my method of risking 1 unit for all rather than flat betting.

 

For the formula of the average moneylines you've played--  that's just all the moneylines added together, divided by the number of games, added to 100.

And I didn't mean double the unit size for both methods (or on both sides of the equation).  I meant you would have to increase the unit size for flat betting until you had risked the same amount overall as with your method.  Right now, flat betting would have risked less.  If you increase it so that you risked the same amount with both methods, the results should be interesting, and less damning of flat betting...

 

I'm going to sleep... good discussing  

 

ps.  playing any dogs tomorrow? :)

 

Hey man, I think you got the wrong idea... what I'm talking about is how much you should bet on any given pick, not what that given pick actually is.  (but I'm playing Houston... and I'm from Philly. Eaton's shaky)

 

Peace

For the formula of the average moneylines you've played--  that's just all the moneylines added together, divided by the number of games, added to 100.

Hold on -- don't go to sleep yet!   .... What do you mean by "added to 100"?

I meant you would have to increase the unit size for flat betting until you had risked the same amount overall as with your method.  Right now, flat betting would have risked less.

Oh OK, I get it. Well I'm not sure how I would uniformly add enough to each bet to equal what I risked doing it my way.

 

Sorry, my mistake.  The formula is more complicated than that...  I think you'd have to average all ML favorites and dogs separately, then give each a weight depending on how many favorites or dogs were played, then multiply and reaverage... I'll figure it out tomorrow and post it.

 

And yea, the second part is pretty much impossible to do, I have no idea how to work that.

Lawstyle, thanks. I would greatly appreciate you doing that for me. If you can't get it done, just let me know and I will see what i can do.

Sorry for keeping you up. Get some rest and I will see you back in here sometime tomorrow.

 

Lawstyle, thank you very much for breaking it down for me, that helped a lot.

thanks

cd

Dang, you're the man. Thanks. You don't have to stay up for it, but I'm going to crank out the numbers within the next 15 minutes. I hope I don't get something wacky.

OK, so here's the verdict Part. 2; I think ...

late 2007 wagers: 262 wagers w/ an average of -10.068
This makes my average moneyline -110, correct?

2008 wagers: 53 wagers w/ an average of -4.063
This makes my average moneyline -104 correct?

OK, unless I misinterpreted your formula, I am, on average betting slight favs. which refutes the premise of this thread. If I messed up and my average moneylines are +110 and +104, then I suppose your argument has some legitimacy but with a heavy disclaimer that if someone leans even a little toward dogs, he is better off steering clear of flat betting.

But I can't see how this is the case. Judging by your method to determine moneyline, lawstyle, I am definitely betting slight favs overall. I just doublechecked your math on this and it makes sense and my numbers come out slightly on the fav side. No offense, but this totally implodes the theory of flat betting. I want you you to prove me wrong, but that's where I'm at right now.

Lantasis - you're missing the point altogether I think.  It is POSSIBLE to win more money betting with a different strategy, that is obvious.

All Vanzack is saying is that IF YOU WANT TO FLAT BET, then don't "bet to win" 1 unit.  He is trying to help people with the proper formula for flatbetting.

If we move this to another sport, say basketball, and let's say you're getting -110 odds most of the time.  In a scenario where you go 2-2 or 3-3 on a day, you're breaking even instead of losing .2 or .3 units.

So, for those that want to flat bet, these equations help. 

How am I missing the point? First of all, I am not talking about always betting to win 1 unit. I am talking about "always risking 1 unit" vs. "flat betting as described in this thread."

The flat bet supporters have stated it is a more profitable method than the alternative unless you are more successful at betting dogs than favs. So then I ran the test on my own results and found that the opposite is true as well. I bet, on average, slight favs and I make less money using flat betting rather than always risking 1 unit. So if flat betting is not good for dogs and now also (as I have proven) not good for favs, it is always less profitable than the alternative I have suggested.

If I've done something wrong, I would like to know, but you have not pointed any such thing out.

you would never bet full kelly, unless your determined probability is 100%, like toss of a coin or roll of a dice.

And then it only determines how many bets you have in relationship to your bankroll, variance and confidence in your apporach which will determine your kelly fraction. Mine range from 1/8th to 1/20th

Sweet, you're back!

OK, I know what you are saying. I understood this premise explained a different way a number of times above. There came a point and time in this thread where I assumed you and Vanzack were correct and was going to complete my calculations merely to confirm your statements.

But while your logic makes sense, it is not hashed out in practical reality, which for me, is my betting log. You would suggest that 315 wagers is a decently sized mathematical model, no? Of course, I would feel better if it was 3,115 instead, but this is what we have to work with. You would also say that my sample size eliminates much of the randomness from a mock experiment of 4 wagers, no? Assuming yes to these questions, then the math of my experiment is more valid than your example, no offense.

I don't know, can you both be correct in theory while I am simultaneously correct in application? That's what it seems like we have here.

I'm just wondering how you justify saying unequivocally that flat betting is more profitable betting more favs than dogs because I do just this but it is less profitable.

Also, remember when you mentioned I am probably risking more units doing it my way than w/ flat betting? Well, I probably agree but this makes flat betting sound very hypocritical. The formulas you and Vanzack gave for flat betting were adjusted to work with a standard betting size of 1 unit and you argued that using this formula = more profits long term than always risking 1 unit instead. I have proven this is not necessarily true.

Maybe this is an idea: Is there a way to tweak w/ the basic flatbetting formula to add more units risked per play to compensate for the disparity?

Anyway, the bottom line is (and my head is starting to hurt from all this): Which method gains you more profits long term? And the answer (for the model I have given, which is decent sized and leans toward favorites) is that risking 1 unit is more profitable than flat betting.

the first thing you have learn is the concept of value. Who cares if the line is -500? If the true price is -600 its a bet. If you can get the probability of the game correct more times then the market you shouldn't flat betting, but using kelly.

I know a great number of sharps and they only bet dogs

I know a great number of sharps and they only bet dogs

A) I would argue they are in the minority for sharps overall.
B) I would argue their capping is lacking since they cannot spot the value in lines where favorites are being undervalued.