MLB Totals Strategy/Research

[englishmike]

When I first started betting on MLB in the UK about ten-years-ago I often found myself taking the over on the lowest total on the board because I really had no clue about pitchers, bullpens ERA, etc.

Consequently, as the years have gone on I've found it very hard to get out of the habit of looking at the lowest totals of the day and blindly taking the over.

I finally decided to track the numbers this year, so what I've done is taken the opening line from SBR Odds on any total of 7.5 and lower and taken the over for one unit.

So far this year the system is +8.30 Units and I shall update the record as the season goes on to see if there is any noticeable advantage in blindly taking the over on any total of 7.5 and lower.

Obviously I accept this is a very small sample size and I do intend to carry update this record reguarly over the next five seasons.

[RickySteve]

Historically you'd have made more money betting under low totals, not over. This season has been unexpectedly high-scoring due mainly to good weather and randomness.

[englishmike]

Historically you'd have made more money betting under low totals, not over. This season has been unexpectedly high-scoring due mainly to good weather and randomness.

Interesting. Do you know of anywhere where I could find five years worth of data regarding totals?

[MonkeyF0cker]

"

[englishmike]

"

[MonkeyF0cker]

Since 2006...

[englishmike]

Thank you very much, that's exactly what I wanted, and I'm encouraged by the data.

Since 2006 any total of 7.5 or less has gone over 714 times and been under 692 times, yielding a unit profit of 22 for the over. 22 Units in three years isn't fantastic and is also low enough it could be argued it could well end up at .500 after six or ten seasons, but nonetheless, 22 units is 22 units.

It's also interesting that, if you took the over on every game of 8 and below you'd actually be 92 units up.

I realise they're not +ev odds so the units would obviously be less.

[MonkeyF0cker]

It's not +22 units though. That doesn't account for vig. If you're betting -110 every game, you'd be -47.2 units. With -105, you'd be -12.6.

[englishmike]

It's not +22 units though. That doesn't account for vig.

Youre way ahead of me, I just edited my post when I realised we're not talking +EV.

[MonkeyF0cker]

You're using the wrong terminology, Mike. EV is used as an acronym for expected value. I'm assuming you mean EVEN or positive odds. Remember that the past doesn't necessitate the future. These totals don't have a very high sample size and could be the result of some random walk. And if they are on account of the market undervaluing the over in this case, it certainly doesn't mean that will continue in the future. Be careful.

[englishmike]

I accept your well made points. You're right, when using EV I was refering to the fact I was not including vig.

And yes, obviously I can see there 's a need to be careful, I was just surprised to see the over 8 number so many units ahead, even accounting for vig.

[Ganchrow]

Since 2006...

Monkey: Not that it affects the overall percentages of course, but it would appear you're double counting games.

Expanding on Monkey's data stretching back to 1999:

[englishmike]

Thanks Ganch, that table is exactly the thing I was looking for.

That said, I'm dissapointed because the numbers bear out what Monkey was saying and are remarkably close to .500 and pretty much tell me there's probably no mileage in getting involved in a long term staking plan involving unders or overs.

Thanks for your time though, it's appreciated.

[MonkeyF0cker]

Monkey: Not that it affects the overall percentages of course, but it would appear you're double counting games.

Expanding on Monkey's data stretching back to 1999:

Oops. Yeah I did. Sorry. I pulled them via team rather than boxscore. Should have done it with the latter method... Thanks for noticing that, Ganch.

[MonkeyF0cker]

I see you modified my half ass table too... Wasn't sure how to post one here. Tried it with simple HTML. That way is much easier. LOL.

[dogman]

Using those figures from the table ,at -110 odds one would bet under only on totals of 11 and 12 but with -105 pricing unders would be bet on totals of 7,9.5,10.5,11,12 and 12.5 for +EV.

So would it be +ev if you could bet those numbers at a better vig, for ex bet under 7 if the vig is +105 which better than -105 ?

The only total that would be played at -105 vig would be the 7.5.

[dogman]

Last sentence was suppose to read, if one wanted to bet the OVER 7.5 at -105 or better

[Regul8er]

Taking games OVER 8 or lower are typically priced in the +105 to +120 range. Very rarely will you see chalk in these situations, as the public pounds the under where they see guys like Haren, Billingsley or Peavy on the hump!

[IrishTim]

Monkey: Not that it affects the overall percentages of course, but it would appear you're double counting games.

Expanding on Monkey's data stretching back to 1999:

Nice table and great info. Just a small quibble, I plugged these numbers into one of my own baseball databases and the % for unders came out to be 51.17%, not 50.83% listed here. Wondering if other info not shown was factored into calculating that average or was it just a mistake? Again, thanks for the data.

[Peregrine Stoop]

not sure those tables really help... should weight everything accordingly as it is rarely straight -105/-105 or -110/-110 betting on MLB totals... they are slanted way more often than not

[CaptainPrice]

doesnt look profitable

[IrishTim]

doesnt look profitable

There isn't much that's blindly profitable, but if you look at enough data over enough years, patterns do emerge that can be profitable. For instance, I believe it's Justin who has over 20 years of NCAAF data that says to play any 6 point teasers with home favorites of 7.5-8.5 when the total is under 48.

[durito]

There isn't much that's blindly profitable, but if you look at enough data over enough years, patterns do emerge that can be profitable. For instance, I believe it's Justin who has over 20 years of NCAAF data that says to play any 6 point teasers with home favorites of 7.5-8.5 when the total is under 48.

That is very different.

That should (i haven't looked at this, just a guess) be profitable with an efficient line. Suggesting it may in fact be more profitable going forward, assuming the market is more efficient now.

Comparing that to the discussion here, what might gave been profitable in the past certainly may not be in the future is especially if the market has tightened up.

[IrishTim]

Valid point, I was just saying that looking at past databases can be helpful in certain instances.

[CaptainPrice]

good stuff

Question
where do you get all the records for this research?

or is it just do it for 20 yrs and report back?