Please tell me if my math is flawed.

Over the baseball season, I have devised a few parlay formulas based on run line stats that I am convinced gives me positive EV. I want someone to show me if my logic is flawed, so I have broken it down to its easiest terms. Please show me the flaw in my logic, if there is one.

These are facts that are undisputable:
1) If we were all to make one million random bets at -110, we would on average expect to win approx 48% (the exact number here is unimportant.
2) Dime lines in baseball (-105) will slightly raise our win expectation, say 49% to make the math simple (although this is not accurate.
3) Random betting will always have a long term negative EV, as is true with all casino games (with a very limited exceptions such as poker or counting cards.

OK, with that established, and realizing billion dollar casinos can be built by only have small % edges, it stands to reason if a player could find one +EV bet over the long run that he could with the correct bankroll, make a nice amount of money. I want to know if I am missing something, or is this not a positive long term EV bet.

MLB run lines: Random betting road teams you will hit about 43%. Favorites (road or home) will win about 55% of the time. Heres the important one- road favorites will win by exactly 1 run 11.6% of the time. Here is where I got my info and it is consistant with other stats Ive seen.

To make the math simple, say you find a road favorite at -140. One a dime line, the home dog would be +130. The road fav runline (-1.5) is also approx +130 (sometimes +125 to +135).

So, if I bet the $100 on the road favorite run line to win $130 + my bet back, and also bet the home dog $100 to win $130, I am wagering $200 to win $30. I would win this bet 88.4% of the time.

So EV = .884(30) + .116(-200) = 26.52 - 23.2 = +3.32 per wager

Player edge then = 3.32/200 = .0166 or 1.66% + Edge. And by shopping lines, this edge can easily reach 2% or slightly more.

Say you had an unlimited bankroll and could wager the optimum amount on every baseball game that has a road fav, an estimate is about 3 per day x six months of games) you could play well over 500 games a years with about the same edge as the house would have in a game of blackjack.

Am I wrong here, or is this bet absolutely +EV? If I am flawed, please be specific and tear this idea apart.

The only flaw in your logic is that although road favorites win by exactly 1 run 11.6% of the time, that includes road favorites of -300, -200, etc. You would need to know the % of road winners at -140 to know if this is positive EV.

This is assuming all your numbers are correct, by the way.

Your argument is fallacious, I'm afraid.

Let's take a simple situation: Team A wins two-thirds of its games against Team B. Therefore, a "fair" line would be: Team A -200, team B +200 (ignore the home/away question for the moment). But that isn't the way it will ever work in a casino--the line will be +190 -210, or something similar. So you are fighting an uphill battle against the "juice", which in and of itself could wipe out your prospective 2%ish "Edge". But let's set that consideration aside for the moment.

Here's the REAL problem: the statement (which I will assume is correct) "road favorites win by 1 run 11.6% of the time" refers to THE SUM OF ALL ROAD FAVORITES. It does NOT, for example, refer to "all road favorites that went off at -140". So plugging in the numbers for that (or any other road-fave situation) and using the GENERAL 11.6% chance of success, must be wrong. (In this case, I would expect the percentage of 1-run wins by the road team to be much greater than the overall average, because a team that is a -140 ROAD favorite is usually a much better team than the home team, i.e., they would win more often than the "average" road favorite, and many of those wins would be by 1 run.)

Using your example, let's say that the actual chance of a -140 road favorite winning by 1 run is 20%. This means you lose both bets 20% of the time, and win $30 80% of the time. This results an EV of -$16 for the $200 action; not good. Similarly, if your line is -120 +110, you would need a lot better than an 88.4% chance of winning (a result other than the favorite winning by exactly one run), because your net gain would only be $10 when you did win (you would lose the $100 you had laid on the -1.5 run line, and win the $100 straight bet on the home dog, for a total of $210). But if the line is -120 +110, the road team is only slightly superior to the home team, and thus a one-run win by the favorite--exactly what you don't want--is that much more likely.

So to truly assess whether this is a viable strategy, you would have to get the numbers for how often does a road favorite win by exactly one run when it is a -110, -120, etc. favorite, and plug THAT number into the price you are getting on the home dog (since that price is where you expect to be generating most of your profit). This might be tricky info to obtain, given that there are betting records, and there are MLB records, but no one entity correlates the two. Therefore, you would need to do some slogging, as in, keep at LEAST one whole year's worth of records.

And finally, not to burst your bubble or anything, but this strategy of yours just MIGHT have occured to those who make the lines, and the prices on run-fave lines just MIGHT be adjusted so as to prevent this strategy from working.