[Todd Deville]

April 19th

Tor(-110) over KC, 3.40 Units.

Tor -1.5(185) over KC, 6.03 Units

[jolmscheid]

Hopefully these hit for you Todd...

[Todd Deville]

Ya I'm not worried, as I mentioned at the beginning, the Kelly Criterion format of betting is extremely volatile. I'll worry if I'm -units in September. I still like all my bets so far, they were long-term profitable. But still, I too hope these hit, I'm pretty confident, I wanted to bet many more Units, but I stuck by the Kelly calculator's unit values.

[rgga86]

Im tailing you in this one BOL

[jolmscheid]

Todd...can you explain to me more what the kelly calculator is?? I don't understand how you arrive at what you are wagering on each game...thanks for the help

[Todd Deville]

Im tailing you in this one BOL

Alright, let's make some money.

[Todd Deville]

Todd...can you explain to me more what the kelly calculator is?? I don't understand how you arrive at what you are wagering on each game...thanks for the help

Ya no problem...here's the link to the calculator...http://www.sportsbookreview.com/Betting+Tool...alculator.aspx

So it looks like there's a lot of different things to fill out with the kelly calculator, but really I only care about filling in two of the squares, the win prob.(probability) and the odds. The top 4 squares should always be set to independent events, consecutive series:1, starting bankroll:100%, and Kelly Multiplier:1.

So what you need to do is figure out what you think the % prob. is that your team will win the bet(be it ML,RL or Total), this is of course the hardest part of handicapping. Some people use trends to figure out the probability of results, others have some sort of simulator they use to try to predict the final score of games. Personally, I look at trends I find over very large sample sizes. So none of this team A is 15-3 ATS against team B since 2006. I need a minimum 200 game sample, and preferably closer to 2000 games. If you can find a trend over this large a sample size, you can be much more certain that it's a real trend and not just luck.

Anyway, so for each game on the day's card, I look if I have any trends, if I do about the % probability in the kelly calculator, and the odds given by the book for that game. It then calculates what the optimal % of my bankroll I should bet. Optimal meaning if I bet that amount, in the long run that will be the most profitable. If you want to be a little more conservative, you can change the kelly multiplier to 0.5, it won't be the best %, but it will be a lesser number, meaning less of those huge swings in +/- units.

This way of betting is really only for the long term. So if you plan on consistently betting baseball for 10 years, or if you plan on using this way of betting for every sport during a season for a couple years. The beauty of the system is you constantly compound your bankroll. So if I bet 5 units to win 4 units(4%), then I immediately add that amount to my bankroll. If my next bet is also 5 units it will actually be a larger amount of $ wagered than the first bet because my bankroll has increased. Now let's say I lost that first 5 unit bet, my bankroll decreases, and if I lose that second 5 unit bet, I lost less money because it was the same % bet but of a smaller bankroll (this makes sense in my head as I'm writing it, let me know if I need to be more clear). So the point is, it maximizes your bankroll when you're on a winning streak, and minimizes your losses when you're on a losing streak.

So really the most important, and hardest, part is figuring out what is the % prob. that your bet will win. If you have bad information then the kelly calculator is useless.

Let me know if I answered you clearly enough, and if you have any other questions.

[Todd Deville]

YTD

Kelly:+0.05
Flat:+1.38

[jolmscheid]

Yeah that really makes a lot of sense. Could you let me know how you pick the percentages of games? Thanks

[Todd Deville]

The percentages are all based on the trends I find, I think you can understand I won't give those up. But really anything else I'm open to talking about, which I believe is more than a lot of other people in this forum are open to. People seem to be a little too protective of their information, I believe you can't give up all your tricks, but I think advice should be given up freely.

[Justin7]

April 19th

Tor(-110) over KC, 3.40 Units.

Tor -1.5(185) over KC, 6.03 Units

Todd,

you are risking almost 10% of your bankroll on this set of correlated plays? There are some real wizards of MLB out there, but do you really think this bet sizing is justified?

[tanner40]

The percentages are all based on the trends I find, I think you can understand I won't give those up. But really anything else I'm open to talking about, which I believe is more than a lot of other people in this forum are open to. People seem to be a little too protective of their information, I believe you can't give up all your tricks, but I think advice should be given up freely.

Good thread Todd. I understand the Kelly calculator but do not use it myself. My question for you is what resources do you have/use that allow you to track historical trends up to 2000 games? Most handicapping sites I have seen well give you a few years of historical information, or the team/pitchers last 10. But up to 2000?

[Todd Deville]

Remember when you look at trends, the higher the sample of games, the more of a TRUE trend it actually is. So if you see team A is 15-3 ATS against team B since 2006, then team A wins 83% of the time according to that stat. If you put 83% in the kelly calculator, no matter what the vegas odds, it's gonna say you should bet a huge %. So you have to think in the long run(thousands of games), would team A always win 83% of the time? Maybe, but you can't have any idea of that from a 15-3 stat. So the most important thing is finding good percentages over large samples.

Finding trends for the league in general, like all Dogs or something like that, will give you larger samples of games. The drawback is these trends are less accurate to the specific teams. Once you add more variables, such as a specific team, at home, as a Favorite, at night, and in July, the sample size will get a lot smaller. So this is where the work of handicapping comes in, there is no easy way of making money in this world, handicapping is no different, you have to put the work in. That work can either be put a ton of hours making models, searching trends, and testing your methods, or those hours can be put into finding a solid group of handicappers who you will follow. It all depends on what you're most comfortable with.

[Todd Deville]

Good thread Todd. I understand the Kelly calculator but do not use it myself. My question for you is what resources do you have/use that allow you to track historical trends up to 2000 games? Most handicapping sites I have seen well give you a few years of historical information, or the team/pitchers last 10. But up to 2000?

It's pretty easy to find trends over 2000 games, but these trends are not specific. It's just looking at Dogs, Favorites, and really general things that all games have. If you just take the last 2-3 years, there were thousands of games played. But, as you said, it's harder to find a lot of games for very specific variables, so I set a minimum for myself of 200 games. Anything less I won't use.

[Todd Deville]

Todd, you are risking almost 10% of your bankroll on this set of correlated plays? There are some real wizards of MLB out there, but do you really think this bet sizing is justified?

I definitely see your point, but there are many handicappers who bet 2-3% on a game and bet 4-5 games, so it would come to the same thing. Some say taking more games spreads the risk, but I'd rather bet few games that I KNOW will be long-term profitable. I have no minimum or maximum amount of bets per day, if I find great value I'll bet it, if I don't I won't bet anything. Back to your 10% of my bankroll question; it is a lot if you are not sure about your trends, also if you cannot handle the heavy swings. I look at sports investing just like the stock market, if you find a profitable trend, in the long run, just like the stock market, you will always have a positive ROI. And just like the stock market there will be good years, bad years, great years, terrible years, if you're smart, in the long run you will always end up on top. Just like in the stock market, you can be conservative or more risky, so if you're starting to find your own trends I would suggest using the flat bet method, and then on the side track your own bets virtually using the kelly calculator.

[Todd Deville]

Justin,

I just realized that my last reply was to you. I thought jolmscheid was the one who had commented about the 10%, that's why I was mentioning the things about if you're just starting out.

I'm sure you are no beginner. So to answer your question simply, if you trust your trends and the results they give you, you input that information into the kelly calculator and bet whatever % it shows as the optimal bet...I trust my trends completely. Although it is fun to bet these games, the point of all the work I put in is to make a lot more money than just a little pocket cash. The kelly calculator might look extreme, but if you have the right outcome probabilities it will payout a MUCH higher ROI in the long run than just flat betting.

Again, I caution people new to this, don't just jump into it. It takes a strong heart, especially in baseball, where you can lose more than 50% of the time and still make a nice profit in the end. If betting is just a pastime, find good trends and flat bet them, it will be much less stressful, and still nicely profitable.

[Todd Deville]

April 20th

Mil -1.5(+167) over Pit, 8.87 Units.

[Todd Deville]

Tor -1.5(+150) over KC, 3.33 Units.

[Todd Deville]

ChiW(+100) over TB, 5.00 Units.

[Todd Deville]

YTD

Kelly:+16.53 Units
Flat:+3.05 Units

[tanner40]

[Todd Deville]

April 21:

Tor(-104) over KC, 8.20 Units
Tor -1.5(+180) over KC, 3.56 Units
ChiW(-107) over TB, 8.24 Units
ChiW -1.5(+180) over TB, 2.00 Units

[PingPong]

with you on white sox, gl

[Todd Deville]

One more for today...

Hou(+137) over Flo, 3.12 Units

[Todd Deville]

YTD:

Kelly:-1.2 Units
Flat: +0.42 Units

[Todd Deville]

April 22:

Was(+162) over Col, 3.56 Units
TB(+105) over ChiW, 6.06 Units

[Todd Deville]

and...

Under 9(-125) Cle @ Min, 5.66 Units

[PingPong]

already given up?