Underdog System - MLB Novice

OLGC_Slayer · 04-15-08, 12:44 AM

Welcome and looking forward to seeing your system in action.

austintx05 · 04-15-08, 12:47 AM

so you're fading the public?

diogee · 04-15-08, 12:51 AM

Best of luck....might be a good idea for you to backtest or to test on paper a little longer. If not I would play with small money until it proves itself.

austintx05 · 04-15-08, 12:54 AM

no offense, but stop while you are ahead. There is no golden system.

Arnold · 04-15-08, 12:59 AM

Originally posted by wallym

Looking forward to sharing the picks and learning from everyone here. Hopefully we can put a set of effective filters to avoid "false picks"...

In order to avoid "false picks", we need to know how your system works, etc. We have no idea why you pick one team over another, so how can we judge your thought/system process?

The most efficient way to test most systems is to go through a database of games. Unless you don't have one, I don't know why waste time on future games.

wallym · 04-15-08, 01:11 AM

The system takes into consideration the % the team is playing for and the current live odds. I think it will be difficult to backtest this...

wallym · 04-15-08, 05:12 AM

15-Apr Picks: (Team / ML Odds / RL Odds)

Minnesota Twins 2.270 1.625
Pittsburgh Pirates 2.470 1.667
Los Angeles Angels 2.040 1.513
Kansas City Royals 2.120 1.526

Arnold · 04-15-08, 12:03 PM

Originally posted by wallym

The system takes into consideration the % the team is playing for and the current live odds. I think it will be difficult to backtest this...

What % are you talking about?

wallym · 04-15-08, 07:33 PM

It's the % shown in the current standings.

curious · 04-15-08, 07:46 PM

Originally posted by wallym

It's the % shown in the current standings.

What does that mean?

Dude if you want help you have to explain how the system works. I have no idea what "the % shown in the current standings" means.

diogee · 04-15-08, 07:48 PM

I would assume winning %.

curious · 04-15-08, 07:51 PM

Originally posted by diogee

I would assume winning %.

Well that clears it up. LOL But what is the criteria? A winning % of what for the dog playing a fav of what winning %?

I'm not fussing at you diogee.

I've seen dog systems where if the dog is at or above .500 and playing at home they are a good bet.

That isn't what this dude is doing.

diogee · 04-15-08, 07:55 PM

Think I got it solved....anytime the team with the better winning % is the dog you play them.

curious · 04-15-08, 08:00 PM

Originally posted by diogee

Think I got it solved....anytime the team with the better winning % is the dog you play them.

Hmmm, that sounds interesting.

What filters would help with something like that?

Weeding out two teams with very low win %? Say Detroit was playing Washington. I wouldn't want to bet on either of those teams. LOL

diogee · 04-15-08, 08:17 PM

Not sure...one possible filter could be the average winning % of their opponents...if the team fits the system but their opponents winning % is .400 compared to the favorites opponents with a winning % of .600 it would obviously make the dog a little bit weaker play depending on how bad the losses were and bullpen stats.

wallym · 04-15-08, 08:20 PM

Thanks for all the posts and my apologies if the system is sounding too vague. The idea behind is converting that % into odds and comparing it with the live odds. It's like "true odds" vs "live odds". If the difference is more than a particular level then that's the pick.

It's totally "mechanical" or "numbers based" way of choosing. The filter I'm hoping for will be the fundamentals side of it -- starting pitchers, streaks, home/away, etc.

diogee · 04-15-08, 08:24 PM

Ahh Ok...filters definitely don't hurt. Do your thing and best of luck to you.

wallym · 04-15-08, 09:02 PM

Thanks Dio...Appreciate it...It's the fundamental things that I'm totally way out...and hopefully will pick up some knowledge around here...

Arnold · 04-16-08, 01:19 AM

Originally posted by wallym

Thanks for all the posts and my apologies if the system is sounding too vague. The idea behind is converting that % into odds and comparing it with the live odds. It's like "true odds" vs "live odds". If the difference is more than a particular level then that's the pick.

It's totally "mechanical" or "numbers based" way of choosing. The filter I'm hoping for will be the fundamentals side of it -- starting pitchers, streaks, home/away, etc.

OK, that's not that hard to back test. Here are some results since 1997 (regular season only):

Betting all road underdogs of >= 130, gets you +79 units.

Betting road underdogs of >= 130, under this system, where the difference between dog's winning % and its odd is >= .05, gets you +151 units.

It also showed most profit per bet when the underdog is exactly on a 1 game winning streak and the home favorite is on a winning streak (.096). If the home team is on a losing streak, then profit per bet is less (.053), but is still greater than having the dog team on a losing streak (.016). If the dog is on a 1 game winning streak, the unit profit per bet is .061.

PS. I need to add that road dogs on a winning streak greater than 1, and home faves on a losing streak, lose you money (-6.5 units). If home team is on a winning streak in the same situation, then you gain +18 units and .059 units per bet.

vassman86 · 04-16-08, 01:45 AM

Originally posted by Arnold

OK, that's not that hard to back test. Here are some results since 1997 (regular season only):

Betting all road underdogs of >= 130, gets you +79 units.

Betting road underdogs of >= 130, under this system, where the difference between dog's winning % and its odd is >= .05, gets you +151 units.

It also showed most profit per bet when the underdog is exactly on a 1 game winning streak and the home favorite is on a winning streak (.096). If the home team is on a losing streak, then profit per bet is less (.053), but is still greater than having the dog team on a losing streak (.016). If the dog is on a 1 game winning streak, the unit profit per bet is .061.

PS. I need to add that road dogs on a winning streak greater than 1, and home faves on a losing streak, lose you money (-6.5 units). If home team is on a winning streak in the same situation, then you gain +18 units and .059 units per bet.

i'm also new to baseball...that is one heck of a database! way to go!

wallym · 04-16-08, 02:00 AM

Thanks Arnold for the info. This are the type of fundamentals info that can be built-in as filters....

wallym · 04-16-08, 02:01 AM

Originally posted by wallym

15-Apr Picks: (Team / ML Odds / RL Odds)

Minnesota Twins 2.270 1.625
Pittsburgh Pirates 2.470 1.667
Los Angeles Angels 2.040 1.513
Kansas City Royals 2.120 1.526

Results:
1-3 for ML
2-2 for RL

wallym · 04-16-08, 05:06 AM

16-Apr Picks: (ML / RL odds)

Pittsburgh 2.82 / 1.76
Florida 2.31 / 1.82

curious · 04-16-08, 01:32 PM

Originally posted by Arnold

OK, that's not that hard to back test. Here are some results since 1997 (regular season only):

Betting all road underdogs of >= 130, gets you +79 units.

Betting road underdogs of >= 130, under this system, where the difference between dog's winning % and its odd is >= .05, gets you +151 units.

It also showed most profit per bet when the underdog is exactly on a 1 game winning streak and the home favorite is on a winning streak (.096). If the home team is on a losing streak, then profit per bet is less (.053), but is still greater than having the dog team on a losing streak (.016). If the dog is on a 1 game winning streak, the unit profit per bet is .061.

PS. I need to add that road dogs on a winning streak greater than 1, and home faves on a losing streak, lose you money (-6.5 units). If home team is on a winning streak in the same situation, then you gain +18 units and .059 units per bet.

Can this be summarized by these rules?:
Ratio of Win % of dog to win % of fav > dog odds
Dog is playing on the road
Dog in on a winning streak
Home fav is on a winning streak

Arnold · 04-16-08, 02:52 PM

Originally posted by curious

Can this be summarized by these rules?:
Ratio of Win % of dog to win % of fav > dog odds
Dog is playing on the road
Dog in on a winning streak
Home fav is on a winning streak

All correct except "Ratio of Win % of dog to win % of fav > dog odds"

I looked only at dog's Win % and the odds on them. So if a dog is .450 (Win %), and the odds to break-even are .4 or less (the difference had to be .05 or more), then it's a play.

Dogs can also be on a losing streak, because it showed profit as well. The only situation that should be avoided is when dogs won more than 1 game in a row, and home fave is on a losing streak.

wallym · 04-17-08, 01:49 AM

Originally posted by wallym

16-Apr Picks: (ML / RL odds)

Pittsburgh 2.82 / 1.76
Florida 2.31 / 1.82

Results:

1-1 ML
1-1 RL

Since 8Apr:

Total Bets: 28
ML Strike Rate: 46.43%
RL Strike Rate: 57.14%

tab · 04-17-08, 05:35 AM

I have a new way of fading out favorites like Peavy who is pitching tommorrow. I am fading out Peavy for a loss. Peavy has thrown 100 plus pitches in 3 straight starts (116, 107 and 100). Eventually a good pitcher like Peavy will get tired and not have his best stuff. It's better to fade out a pitcher who has won 3 straight starts and where the pitcher threw over 100 pitches in each of those starts.

wallym · 04-17-08, 07:47 AM

17-Apr Picks: (ML / RL Odds)

Florida 2.330 / 1.800 (home - streak: W2)
Washington 2.470 / 1.690 (away - streak: L2)

Arnold · 04-17-08, 11:57 AM

Originally posted by tab

I have a new way of fading out favorites like Peavy who is pitching tommorrow. I am fading out Peavy for a loss. Peavy has thrown 100 plus pitches in 3 straight starts (116, 107 and 100). Eventually a good pitcher like Peavy will get tired and not have his best stuff. It's better to fade out a pitcher who has won 3 straight starts and where the pitcher threw over 100 pitches in each of those starts.

Why?

All good pitchers throw 100+ pitches a game. Pitchers who throw less, usually have bad games and are taken out early.

wallym · 04-18-08, 06:53 AM

Originally posted by wallym

17-Apr Picks: (ML / RL Odds)

Florida 2.330 / 1.800 (home - streak: W2)
Washington 2.470 / 1.690 (away - streak: L2)

Results:

0-2 ML
1-2 RL

Total Bets: 30 (mechanical picks -- no filters)
ML Strike Rate: 43.33%
RL Strike Rate: 56.67%

wallym · 04-18-08, 07:07 AM

18-Apr Picks: (ML / RL Odds)

Washington 2.150/1.513 (away / streaks: L1 @ L3)

Not too many selections lately...the live odds vs real odds are very close...

wallym · 04-19-08, 09:20 AM

Originally posted by wallym

18-Apr Picks: (ML / RL Odds)

Washington 2.150/1.513 (away / streaks: L1 @ L3)

Result:

1-0 ML
1-0 RL

Total Bets: 31
ML Strike Rate : 45.16%
RL Strike Rate : 58.06%

wallym · 04-19-08, 09:21 AM

19-Apr Picks: (ML/RL Odds)

Washington 1.980 / 1.455 (away / streaks: W1 @ L2)
San Francisco 2.080 / 1.455 (away / streaks: L3 @ W1)

wallym · 04-20-08, 06:00 AM

Originally posted by wallym

19-Apr Picks: (ML/RL Odds)

Washington 1.980 / 1.455 (away / streaks: W1 @ L2)
San Francisco 2.080 / 1.455 (away / streaks: L3 @ W1)

Results:

ML : 1-1
RL : 2-0

Since 8Apr:
ML: 15-18 (45.45%)
RL: 20-13 (60.61%)