[St. Andrew]

Hitting 55% in my season thread, which is okay I guess, but I am constantly tweaking my strategy to get and edge. Here's what I am trying today to see how it works:

1. Look at each game and determine the mean total runs given up by each team when that pitcher is on the mound. Obviously the bullpen is built into this. The pitcher must have a minimum of three starts.

2. Then, determine the mean runs scored by each team over the last 10 games. Obviously recent injuries will be built into this.

3. Then, take the average of the runs given up (#1) and the opponent's mean runs scored (#2), round to the nearest run, and get a projected game score.

To make a play on a side or total, the projected system total or side must differentiate from the line by at least 2 runs. However, make a play on a difference of 1 run IF the payoff is in positive money. If there is a difference of three runs, use that difference in a parlay.

Here are today's projected scores and plays:

Astros 5
Dodgers 6
Play: total over 8.5

Rockies 5
Padres 4
Play: Rockies SU +107

A's 4
Rangers 4
Play: total under 10

Braves 6
Pirates 6
Play: total over 9.5

Reds 4
Mets 7
Parlay: total over 8 / Mets SU
+170

DISCLAIMER: before all you statisticians weigh-in and say "your sample size is insignificant", let me say "I KNOW"!

However, I think there is enough utility in it now to begin using it. Will it fluctuate around the bell curve early and be erratic? Absolutely! However, I think it makes as much sense to use this as any other current capping system. I think, longterm, this will hit at greater than 60%.

$200 on each play

Good LUCK to all today, no matter how you're fighting the book!:

[OLGC_Slayer]

Interesting. Are you going to post daily plays using this system Saint?

[St. Andrew]

Originally Posted by OLGC_Slayer

Interesting. Are you going to post daily plays using this system Saint?

I'll do this for at least three days in this thread. We'll see how it goes and take it from there.

[St. Andrew]

Originally Posted by St. Andrew

Reds 4
Mets 7
Parlay: total over 8 / Mets SU
+170

This refers to Game #1 (Santana/Belisle).

[MrX]

Originally Posted by St. Andrew

3. Then, take the average of the runs given up (#1) and the opponent's mean runs scored (#2), round to the nearest run, and get a projected game score.

This step is fatally flawed. Say league average runs scored is 4.5/game. Now say you have team A that scores 6 runs/game playing team B that allows 5.5 runs/game. You're method predicts 5.75 runs.

This is obviously not right. Team A scores 6 versus average pitching. Now they are facing sub-par pitching and they will score less?

What you want to do is something more like this: Expected runs = RS*(RA/LA) = 6*(5.5/4.5) = 7.33

I will bite my tongue on the sample size issues as per your request.

[St. Andrew]

Originally Posted by MrX

This step is fatally flawed. Say league average runs scored is 4.5/game. Now say you have team A that scores 6 runs/game playing team B that allows 5.5 runs/game. You're method predicts 5.75 runs.

This is obviously not right. Team A scores 6 versus average pitching. Now they are facing sub-par pitching and they will score less?

What you want to do is something more like this: Expected runs = RS*(RA/LA) = 6*(5.5/4.5) = 7.33

I will bite my tongue on the sample size issues as per your request.

I think you may be misunderstanding Step #2. When I say "mean runs scored by each team" I am referring to the two teams in the matchup, not the league averages. Now maybe I am misunderstanding your response.

Either way, I'll see how this plays out and abandon it if (when) it fails after minimum three days. Thanks

[St. Andrew]

"What you want to do is something more like this: Expected runs = RS*(RA/LA) = 6*(5.5/4.5) = 7.33"

This makes sense as long as I could consistently get a running total of league avg runs on a daily basis. I'm not sure what site would provide that.

[MrX]

Originally Posted by St. Andrew

I think you may be misunderstanding Step #2. When I say "mean runs scored by each team" I am referring to the two teams in the matchup, not the league averages. Now maybe I am misunderstanding your response.

No, I understand what you're doing, but your method for calculating the expected runs just doesn't work. Take another example. Team A is an offensive juggernaut that scores 8 runs/game. Team B has average pitching of 4.5 runs/game.

Since team A scores 8 runs/game versus average opponents and team B has average pitching, the expected runs for the matchup should be 8 runs. Your method will predict 6.25. That's a big difference.

Originally Posted by St. Andrew

Either way, I'll see how this plays out and abandon it if (when) it fails after minimum three days. Thanks

Carry on, then, but three days will tell you nothing meaningful.

[element1286]

I did something like this for a while, but I used three years stats. And I didn't use anyone who didn't have 60 starts in three years. As far as teams stats, I didn't really take those into account at all, always gave me weird totals like Mr X was saying. It was ok a predicting totals, there was rarely more than one play per night. I think your use of this years stats is flawed.

[St. Andrew]

Originally Posted by MrX

No, I understand what you're doing, but your method for calculating the expected runs just doesn't work. Take another example. Team A is an offensive juggernaut that scores 8 runs/game. Team B has average pitching of 4.5 runs/game.

Since team A scores 8 runs/game versus average opponents and team B has average pitching, the expected runs for the matchup should be 8 runs. Your method will predict 6.25. That's a big difference.



Carry on, then, but three days will tell you nothing meaningful.

In this example, realize what you just said: my system will predict Team A to score 6.25 runs. That is almost two full runs more than Team B gives up --on average-- per game, as it should: Team A is an offensive juggernaut. I'm not saying that you don't have a valid point to your argument, but your argument is based on mathematical principles that are only relevant in a model of 100,000 or so wagers, not in a season of only 150 or so games that I will play.

So this where you say: "then why even bother doing this if you know it's flawed and is not based on statistical relevance"?

All of baseball betting is flawed and filled with human dimensions. However, I "instinctually" feel that this system will tend toward the means I have discussed, even if these means are not arrived at using sound statistical laws. If that sounds ridiculous, then it is because I am a gambler and not a statistician.

Either way, whether my early faith in this "system" is flawed or not, it gave me the confidence (misguided confidence no doubt) to bet a parlay that is about to cash.

[MrX]

Okay, just trying to help.

Have fun.

[St. Andrew]

Day 1 system goes 1-3 and a push, but the win was the parlay.

Net loss of $260 on $200 wagers

Weird things today....Dodgers getting shut-out at home by Sampson? Braves scoring two runs in the first inning, then none for the rest of the game? I'm not sure if anything can predict these scenarios.


I'm committed to this for two more days. Day two coming up.....

[austintx05]

blah blah blah

[Smutbucket]

I am looking forward to the system plays and am tweaking some systems myself...Sounds interesting and makes sense to able and should get a least 55% ...i will follow please continue to post despite criticism and offset of 1 day...iam following a lot of systems and try to tweak out and record the best ones and weed out the bad ones. thanks

-smut

[St. Andrew]

Originally Posted by austintx05

blah blah blah

Sunday System Plays

Marlins +103
Cards +105
Red Sox / Twins over 9

[Clueless_Norway]

Good Luck.... Im on the Marlins and Cardinals (Without Isringhausen now so..)

[St. Andrew]

System goes 2-1 on Sunday (+$206)

Net gain/loss overall on $200 wagers: -$46

Monday Plays in a few....