Let's now move a few decades into the present.  Recently, we had a year where 303 different pitchers started a major-league game. 303 different guys.  Is that amazing, or what?  It is to me.  But here's an even more amazing stat.  Of those 303 pitchers, exactly 2 pitchers had more than a single complete game for the whole season. 2 out of 303. Wow!  And those two happened to be former Cy Younger's Roy Halladay of the Blue Jays and Brandon Webb of the DBacks.

So today's starter has no intention or desire or even a remote probability of finishing that game for you - whether he's had an excellent effort or not - and that's why you often see guys toss 6 good - but not great - innings but their team loses because, guess what, there are still 3 more innings to get the other guys out.  It's very, very dangerous these days to rely solely on starting pitcher stats when you bet baseball - for these and other reasons.

And Santana?  The guy who was the highest-paid pitcher in baseball history? With a $137.5 million dollar contract over 6 years?  That sounds more like a Mega Millions Lotto jackpot than a salary, eh?  Well, if you had bet the Mighty Johann in every one of his 28 starts during the 2007 season and you wanted to win $1000 on every game (risking the favorite price to win that dime - and risking the dime to win any dog price he might have been) - what would have happened to your bankroll?  You would have ended up the season a cool -$11,100 on this phenom.  Down more than 11 grand.  How about them apples?  Still think there was value in that -3.00 play with Santana?  Maybe yes...maybe no...but there's no maybe about this fact: Bob Akmens won't give you a crap-value play like that, folks.

Since 1978, Bob Akmens Sports has finished #1 in more different categories of documented baseball handicapping contests than any service ever - dead or alive. You'll find an up-dated total of #1-finishes further down the page.  We're also the defending national handicapping champs by virtue of finishing #1 in the 2008 MLB Post-Season Contest run by the Las Vegas Sports Monitor (see our award-certificate here).  And in 2007, we were #1 in both The Sports Eye (TSE) and The Sports Monitor (TSM) contests during the entire 2340-game 2007 regular-season (yup – that’s how many games there are scheduled in a baseball regular-season.)  In fact, we finished in first-place in an unprecedented 6 different categories in the TSM contest - the most ever in the 20-year history of a TSM baseball competition. You can find a complete list of those #1's here.

I'll share something that's not-so-secret and something my long-time clients already know: of all the sports to watch, I love baseball most.  And that's for a lot of reasons.  You don't have the mindless violence you see all too often in hockey games, or the other contact sports of football and basketball.  And, in spite of all that spitting going on, baseball really is more of a thinking-man's game than the other sports and strategizing plays a big role in a game's outcome - and in handicapping it.

In recent years, I've personally developed sophisticated computer models to come up with my baseball plays.  To do this, I started by inputting literally hundreds and hundreds of possible game-variables and seeing how they affected a game's outcome.  Things like how does a pitcher's team do in his road starts as opposed to his home starts. Then another level deeper you analyze how that pitcher's team does in his road starts when he comes off of a win - or a loss.  Then yet another level deeper and you analyze how his team does in his road starts, after a loss, with his regular days of rest - and how they do with less, or more, rest.  And so on - and so on. You should get the picture.

And after looking at so many variables - and their consequent results - I boil down the whole brew to the dozen-or-so most predictive variables: the ones that are most important in the winning - and losing - of baseball games.  In the next step to creating a predictive model, I weigh each of those variables that made the final cut and assign it a relative percentage-importance in that model.  Continuing the model-building-process, I test that model - I test it and I test it and I test it - over and over and over - on thousands of previous games.  This is called back-testing and every predictive model ever created for anything: predicting how a hurricane will track; how stock markets may move; and how baseball games win and lose - needs extensive back-testing.

But I'm not done yet: I still have two more very important steps, one finite and one virtually infinite.  The finite one is that even if the back-testing works to my satisfaction, it's still based on something that's not real-time - it's based on completed games and is ex post facto - after-the-fact. So, I start to apply my model to real-time games - and bet them myself - but I don't yet release those plays until I'm sure this model works in real-time handicapping.  Then - and only then - do I start to release plays based on that 
model.

And that final "infinite" step?  I call it infinite because it never stops.  The model becomes an organic thing - and grows, and changes if necessary, as time goes on.  This is important.  The changes that take place in college and pro sports move at different speeds. Some changes happen right away, as for instance occurred when the NHL put new rules into play a few years back. These changes were designed to do one thing: create more offense in a game which had stagnated to many low-scoring affairs.  And low-scoring affairs are not something the average American sports fan likes.  I believe that's why soccer has never caught on like it has in the rest of the world; we're apparently just too impatient to wait for teams to score.  When the NHL put those rules into play, instead of having a common totals-line of 4.5 goals a given game, you suddenly had a line of 5 or even 5.5 goals on those very same teams when they matched-up. 

Other sports evolve more slowly. As I noted earlier, when I started doing this for a living a starting pitcher's goal in every single game was a complete game shutout, while today that's almost non-existent.  It's as if we've transformed into some alternate reality if you've followed baseball for as many years as I have.

There's no mystery why a guy like Tony LaRussa has not only won with every team he's managed, but has finished in first-place with each of his three teams: the A's, the White Sox and the Cardinals - winning a couple of World Series along the way.  Watch LaRussa in the dugout when they cut to him. He's always checking his stats: of opposing hitters, pitchers, situations, etc.  In the old days, he used to have notebooks he'd look at all the time; now, he checks his stats on a laptop most of the time.  This guy does his homework and understands, as I do as a handicapper, that you have to look at the percentages - the actual reality of what has happened - and not just guess what you should do based on some sort of ancient, outmoded baseball code.

As to the actual bets you make in baseball: is it better to bet sides, or totals? Favorites or dogs? The money-line or run-line?  Well, the short answer to all of those questions is: Yes.  You bet what looks like will be a winner.  I bet what my model tells me to do.  Sometimes, I do so painfully because it might not be the play I'd use if I hadn't crunched the numbers.  But almost always the model will rule the day - and you know what?  I've done much better with my own betting since I pretty much automated my handicapping procedures using these computer-models.  I don't get caught up in my own winning streaks or losing streaks - I just bet what the model tells me to.

I do get questions every baseball season from clients who are interested in betting the run-line instead of the money-line on a game.  If you're not familiar with that, here's how  that works:  let's say Steroid-Boy Clemens was a -1.90 favorite on the money-line.  To win a bet on that line, you'd lay, say, $190 to make a $100 profit. And regardless of  whether Clemens finishes that game - which, as pointed out, virtually no pitcher does these days - all you need is for Clemens' team to win the game for you to win the bet.  Of  course, if his team loses the game, you lose your $190.  Now, what if you could lay only $110 but his team would have to cover a -1.5 run line as well?  Meaning, his team now  has to not only win the game, but they have to win the game by 2 runs or more for you to win your bet.  The bottom line is if you win with this run-line, you still win $100. If you  lose, you lose only $110.

It happens that sometimes a run-line is an interesting adjunct to the money-line, meaning a useful additional bet.  Is it a good substitute for a money-line bet, such that you don't bet the money-line at all?  Not in my estimation.  Here's why:  baseball is, as noted above, a game of strategy.  The objective is to win the game.  The entire nature of  baseball is to work to create that sole run crossing home plate.  Because if you can do that, your pitching staff (and workmanlike defense) can shut down the opponent and you  can win with the sole skinny run.  How often do you see a football final of 7-0?  Hardly ever.

Because the nature of football is to endlessly drive forward on each and every down -  unless, of course, you're trying to run out the clock with a lead.  And sheer momentum sometimes creates scores in football.  Not so in baseball.  So, laying a run-line puts the  game into another dimension - you now have to not only win, but win by multiple-runs.  This is just not something I'm fond of and I just don't give out many run-line plays. What I will do for any client is this: call me or email me with some time to research this for you and I will tell you if there's value in any given run-line you're interested in.

I hope some of these points help you to make you money betting baseball - I know they have made me a better bettor over the years.